A TAXONOMIC STUDY OF THE PHYTOSEIINAE (FAMILY LAELAPTIDAE) PREDACEOUS UPON TETRANYCHIDAE OF ECONOMIC IMPORTANCE

Transcription

1 A TAXONOMIC STUDY OF THE PHYTOSEIINAE (FAMILY LAELAPTIDAE) PREDACEOUS UPON TETRANYCHIDAE OF ECONOMIC IMPORTANCE by H. H. J. NESBITT, M. A., Ph. D. (Toronto), D. Sc. (Leiden) Carleton College, Ottawa, Canada The increased importance which the European red mite (Paratetranychus pilosus (Can. et Fanz.)) (= Metatetranychus ulmi (Koch)) has assumed in recent years has led to an intensive study of its biology and natural history. In the course of these investigations many workers, and in particular those in Nova Scotia (vide Lord, 1949), have become convinced that this pest can be controlled, on apple trees at least, by natural means and that some of the most active agents in its eradication are the representatives of that group of predaceous mites which Vitzthum (1941) placed in the subfamily Phytoseiinae Ber'lese, 1916 ). As the late Dr. A. C. Oudemans of Arnhem 1 included many if not most of these species in the genus Typhlodromus as he conceived it, this paper is in essence a revision of that genus. Presumably because of their small size and limited distribution, which is largely contingent upon readily available populations of their hosts, little attention has been paid to these predators from either the ecological or taxonomic point of view. A cursory survey of the literature pertaining to the predaceous relationship which exists between the Phytoseiinae herein to be discussed and the tetranychid mites may serve as an appraisal of this economically significant group of mites. Koch (1839) in describing what now appears to be a typhlodromid, viz., Gamasus vepallidus, made no reference to its possible predaceous habits. Scheuten (1857) thought that the eriophyids which he found associated in numbers with his Typhlodromus pyri were its offspring. Berlese ( ), however, had a better understanding of these relationships and was able to state in his redescription of G. vepallidus as Seius (Seiulus) vepallidus (K.) that it was a predator of small acari as well as being a mycophage. His countryman, Ribaga (1902), writing of the gamasids associated with plants, failed to note any such relationship. In 1906 the American entomologist Parrott, as a result of his studies on the eriophyids of apple and pear trees, was able to state, "The apple and pear blister mite is much subject to the ravages of a gamasid mite Seius pomi (Parrott). This was very abundant this season upon infested trees and undoubtedly material- 1) For a list of the genera included in this group see p. 4, footnote.

2 2 H. H. J. NESBITT ly assisted in reducing the numbers of blister mites". Six years later Quayle (1912) reported a gamasid preying upon a tetranychid which he called Tetranychus mytilaspidis Riley (? Metatetranychus ulmi (Koch)). In 1914 Ewing described in a rather general manner the different instars of Seius pomi and reported that in Oregon he found it to be one of the most efficient natural enemies of Tetranychus telarius (L.) and of a species which we now believe was probably the European red mite. These findings relative to the predaceous habits of typhlodromids were in the main substantiated by Newcomer and Yothers (1929) working in the apple growing areas of the Pacific north-west (the state of Washington and surrounding areas) and by Gilliatt (1035) in a similar ecological area in Nova Scotia. The latter author was somewhat more outspoken in his claims for Seiulus pomi (Parrott) : "This mite is probably the most important of all the predators attacking the European red mite in Nova Scotian orchards, not that it consumed more mites during the many periods under observation in comparison with other predators, but on account of its being active for the entire season from early spring before the red mite eggs hatch until well after all winter eggs have been deposited". Somewhat similar but less emphatic results have been reported by Kuenen (1945) working with Typhlodromus similis (Koch) ) in x the Netherlands and by Massee (private correspondence) in England. As a result of four seasons of experimental and observational work in the main fruit growing areas of Canada, I am convinced that the complex of species loosely referred to as ''Seius (Seiulus) pomi" is one of the most useful agents in the control of all tetranychid mites. For many years the systematics of the species of the Phytoseiinae has been in a state of confusion. Many factors have conspired to bring this about. None of the older type specimens were available for study and most original descriptions were so vague that they could be variously interpreted. Finally, there was almost a complete lack of liaison between the zoologists in Europe who were engaged in taxonomic pursuits and those in America who were interested in the biology and ecology of the species involved. Some of this trouble has now been remedied. Large collections, numbering hundreds of specimens of some species, of Canadian, American and European material have been compared with the Oudemans specimens in the Rijksmuseum van Natuurlijke Historie, Leiden, and a diligent search has been made of the older literature in an effort to understand what the writers had in mind, or rather hand, when they made their descriptions. While the type specimens are still not available and repeated search leads one to believe that they are gone 1) Berlese regarded this species as being synonymous with Amblyseius obtusus K. There is no question about its inclusion in the genus Amblyseius, but there is some doubt about the specific synonymy.

3 A TAXONOMIC STUDY OF THE PHYTOSEIINAE 3 forever, collections from the type localities have proved an aid in interpreting the original descriptions. Some word of thanks should be extended to Dr. H. Boschma, the Director of the above mentioned museum, who kindly placed the Oudemans collection at my disposal in the summer of 1948 when I was in Leiden and who has given me permission to include many of Oudemans' original drawings in this paper. As first reviewer of the group I have had to exercise a certain amount of choice in the fixing of genotypes and hence the genera, but at all times great care has been taken because I realized that in so doing I was adopting a somewhat arbitrary taxonomic position. It is my belief, however, that the exigencies of the situation warranted such an action and that it is preferable to maintain taxonomic names, which in the past have expressed a true biological relationship rather than to discard them because the type specimens on which they are based are no longer available. In 1929 Oudemans began his studies of the typhlodromid mites by separating the known species into two groups on the nature of the dorsal setae. For those with smooth setae he resurrected Scheuten's 1857 genus Typhlodromus stating that as Scheuten's type pyri (1857) synonymous w a s with Gamasus vepallidus Koch, 1839, the latter would have to be designated the type of the genus. Antithetically the species with the rough, feather-like dorsal setae were relegated to the genus Seiulus Berlese, 1887, type Seiulus hirsutigenus Berl, 1887 (vide infra). The genus Typhlodromus as thus conceived and understood by Oudemans was a composite group including several widely divergent forms ). The Dutch author recognized this and in expressed doubts in his own mind about the advisability of maintaining such a heterogeneous group but stated ) that as he had never seen pyri he did not 2 dare to break up the genus Typhlodromus into the different subgroups which were apparent. He did, however, indicate possible relationships and cited the different species which he would group together (vide 1930 paper). The genus stood as such until June of 1948 when Dr. Philip Garman decided to split it into its component parts by means of such characters as the number and disposition of the dorsal setae and the size and shape of the anal 1) Oudemans never, to my knowledge, apart from designating the type and setting forth the description of several species, gave a formal description of this genus. My knowledge of his conception of the genus is based on (1) his species descriptions; and (2) his type specimens and figures deposited in the Rijksmuseum van Natuurlijke Historie, Leiden, Nederland. 2) Oudemans' words read, "Zoolang ik de op Pirus communis (en wellicht op andere vruchtboomen) voorkomende T. pyri Scheuten 1857 [he apparently overlooked the fact that in 1929 he stated that pyri was a synonym of vepallidus], type van het genus Typhlodromus Scheuten 1857, niet met eigen oogen aanschouwd heb, waag ik het niet de door mij beschrevene soorten in subgenera, of genera, te splitsen".

4 4 H. H. J. NESBITT plate of the female. On the basis of these criteria the genus Typhlodromus was retained for the group of species where there are upwards of twenty-five pairs of dorsal setae and where the anal plate is small and seldom bears more than three setae. Ribaga's 1902 genus Iphidulus (type I. communis Rib., 1902), on the other hand, was revived from obscurity for the species with only seventeen pairs of dorsal setae and with the larger anal plate which bears more than three setae. Two questions must now be proposed: was Garman justified in his separation of the genus Typhlodromus (as understood by Oudemans) into its component parts, and what are the correct generic names to be applied. The first question, I believe, may be answered in the affirmative if investigators in this field are willing to set aside a purely legalistic interpretation of the relationships of species in favour of biological realities. Garman was correct in his assertion that Oudemans' group of species falls quite naturally into two (later discussion will reveal at least four) subgroups but, as I subsequently hope to shew, I do not think the generic names which he used were the correct ones. Before any definite decision can be made concerning Garman's choice of generic names the genera ) of the Phytoseiinae which are involved and those 1 proposed in 1902 by Ribaga will have to be examined. The genus Typhlodromus was established in 1857 by Scheuten for his new species T. pyri. The description and figures which he presented in support of his contention that he had an undescribed mite are not sufficiently detailed to be of much value. In describing his species, however, he stated 1) Vitzthum (1941) page 767. "5. Subf amilia: Phytoseiinae Berlese Genera und Subgenera: 1. Typhlodromus Scheuten 1857 ( Seiulus Berlese 1920), Typus: Gamasus vepallidus C. L. Koch 1839 (= Typhlodromus pyri Scheuten 1857). 2. Seiulus Berlese 1887 (= Echinoseius Ribaga 1902), Typus: Seiulus hirsutigenus Berlese Phytoseius Ribaga 1902, Typus: Gamasus plumifer G. Canestrini et Fanzago Iphidulus Ribaga 1902, Typus: Iphidulus communis Ribaga Amblyseius Berlese 1914, a) Amblyseius s. str., Typus: Seius obtusus Berlese 1889 ex Koch. b) Seiopsis Berlese 1923, Typus: Amblyseius (Seiopsis) brevipilus Berlese Kleemannia Oudemans 1930, Typus: Zercon pavidus C. L. Koch 1839."

5 A TAXONOMIC STUDY OF THE PHYTOSEIINAE 5 that he found it on pear leaves associated with eriophyids and in his drawing, which might otherwise be taken for a generalized Laelaptid, he shewed the outline of the gut. This latter point is worthy of notice because in the usual Laelaptids which are found on apple and pear leaves only those forms *) which feed on red mites and have a weakly sclerotized dorsum shew this character. From this very pertinent observation and from his description I am convinced beyond reasonable doubt that Scheuten had one of the common predators of the European red mite and that he was justified in establishing a monotypic genus for it. The question, however, of the type and, dependent upon that, the limits of the genus sensu stricto, is a controversial point which must be settled. There are two possible solutions to this problem. One is to declare that as there is no holotype or lectotype material available we are justified in treating Typhlodromus as a nomen dubium and erecting a new genus for many of the species which Oudemans placed in his genus Typhlodromus. The other is to attempt to identify Scheuten's species within certain justifiable limits and so fix the genus if not on one species, at least on a group of very closely related forms. The first procedure is a sound taxonomic expedient which avoids entirely the purpose and spirit of modern systematic thought, the second is a somewhat arbitrary practice which, however, preserves the spirit if not the letter of the law and maintains a well-established generic name. Furthermore, in essence the second method of dealing with this problem has been condoned by the Paris meetings of the International Commission on Zoological Nomenclature ). 2 1) Included in this group are the species known and described by Oudemans as Typhlodromus tiliae, rhenanus, tiliarum, and finlandicus, and Seiulus spoofi. I doubt if Scheuten would have confused the last mentioned species with any of the former because its large serrate setae make it quite distinct even under the lower powers of magnification. 2) A pertinent excerpt from the Proceedings of the International Commission on Zoological Nomenclature at its Session held in Paris in July, 1948; 4 (4/6) : 76, 25th May 195JO; reads as follows: "(g) that, where an author, when publishing the name of a nominal species, either (i) omits to specify the material on which that nominal species is based and it is later found impossible to trace that material, or (ii) specifies his type material, but that material either (a) is so imperfect or in such bad condition as to render it impossible to recognise the taxonomic species of which it consists, or (b) was lost or destroyed before the identity of the taxonomic species in question w as established, the following rules are to be applied: T (1) where, in spite of the lack of a holotype or lectotype or, as the case may be, of a recognizable holotype or lectotype, specialists are able to recognise the taxonomic species represented by the nominal species in question the name of that nominal species shall apply to the taxonomic species so recognised." Certain portions of the above excerpt are applicable to the case in hand of Typhlo-

6 6 H. H. J. NESBITT If we adopt the second alternative what then is Typhlodromus pyri Scheuten? From collections which I have made in the Netherlands, Belgium, and France and which have been made for me in and about the neighbourhood of Bonn, Germany, where Scheuten states he found his specimens I am convinced that the species under discussion was either Typhlodromus tiliae Oudms., the almost identical T. tiliarum Oudms., or T. finlandicus (Oudms.). These three species are similar morphologically and ecologically and can only be separated with difficulty. Of the three species just noted I am inclined to favour the first mentioned, viz., Typhlodromus tiliae Oudms., because it is by far the most widespread and common of the predators of the European red mite and eriophyids, because it shews the outline of the gut very clearly after feeding on red or clover mite ) eggs, and 1 because anatomically it corresponds with the figures and descriptions published by Scheuten. Oudemans on the basis of rather limited collections ) but a wide knowledge of the literature took a somewhat different view. In 1929 he was con 2 vinced, for reasons which I have not been able to discover, that Scheuten's type from pear was synonymous with Koch's Gamasus vepallidus 1839 from elm, whilst in the next year he cast doubt on this by stating (1930) that he had not yet seen the type of T. pyri. Subsequently in 1941, Vitzthum either was ready to accept Oudemans' former statement or had reasons of his own for arriving at the same conclusions for in his last great work which appeared in that year he accepted G. vepallidus Koch as the type of the genus Typhlodromus. Before this synonymy can be accepted, however, an exact knowledge of the nature of Gamasus vepallidus should be available. Koch's drawings are too vague to be of value and those published by Berlese in 1897 complicate dromus pyri. Clause (i) is relevant in that Scheuten made no reference in his description to his disposition of his type material, whether it was placed in a museum or in a private collection. All we know is that to all intents and purposes it is lost. In all probability, living when he did, he did not realize that type specimens had any great significance. His writings also, wherein he referred to glycerine preparations, would indicate that he did not make permanent preparations. Clause (ii) b is also helpful because as far as we know no person ever saw the types. Oudemans (1930) stated that he had not seen any and Berlese made no reference to any knowledge of Scheuten's work. 1) From the description and figures given it would appear that one of the other mites, viz., Sannio rubrioculus Scheuten which Scheuten found in his collections from pear and linden was the clover mite Bryobia praetiosa Koch, a frequent prey of Typhlodromus spp. 2) I doubt if Oudemans had in all more than 75 or 80 specimens of Typhlodromid mites. The great bulk of these were submissions by others rather than his own collections. In fact in later life he regretted that he had not been able to spend more time in the field observing some of these forms under natural conditions.

7 A TAXONOMIC STUDY OF THE PHYTOSEIINAE 7 the issue because it appears that he either misidentified the species, published a composite drawing or incorrectly drew what he saw ). Thus we are in the 1 position of being unable on anatomical grounds to identify G. vepallidus any more exactly than to a group of species within the genus Typhlodromus (Oudemans' conception). There are, furthermore, no ecological or natural history notes in this description as there were in the case of T. pyri to aid us in arriving at a satisfactory conclusion. Oudemans (1905) has stated that his Seiulus rhenanus is very similar to G. vepallidus but that it differs in a few minor details, viz., in lacking two coarse setae behind the vertical setae; in that it has only one pair (not two) of coarse setae at the posterior margin of the body; and in that the copulatory organs of the two species are somewhat differently shaped. It is difficult to understand how the Dutch author could make these statements because Koch's drawing of vepallidus looks like a simplified version of almost any one of half a dozen different species and from extant records (both written and microscopic slides) he could have had little knowledge of vepallidus apart from the original description. On the basis of the above argument and in view of the fact that G. vepallidus was found on elm ) whereas T. pyri has been taken from pear, 2 apple and linden, the evidence seems to be in favour of the opinion that pyri and vepallidus can not be synonymous and that pyri should stand as 1) Berlese's figures of the dorsal surface though far from exact (vide A.M.S. fasc. 54, N8. fig. 1) and of the chelicerae (ibidem, figs. 5 and 6) indicate that his vepallidus belongs to that group of species which is typified by T. tiliae Oudms., i.e., those in which there are not more than 17 pairs of setae on the dorsal shield. The ventral view (ibidem, fig. 2), on the other hand, shows a small anal plate which would place the species under discussion in the other group with such forms as T. domesticus Oudms. and T. pomorum Oudms. In interpreting these figures and the accompanying descriptions we can assume: i) that Berlese correctly drew what he saw which would indicate that he had a species which has subsequently never been seen by either Oudemans or later workers; ii) that he confused two species, drawing the dorsal surface of one and the ventral surface of another, which would be almost impossible because the species comprising the two subgroups are seldom if ever found together in the same habitat; iii) that he incorrectly drew either the dorsal or ventral surface. I am inclined to the third explanation and to adopt the attitude that the ventral surface of the female (ibidem, fig. 2) was incorrectly drawn because, as has already been stated, the figures of the dorsal surface and the chelicerae are applicable within justifiable limits to the tiliae group of species, whereas the anal plate, whilst typical of the other group might have appeared as a small oval platelet rather than the full hexagon under certain methods of clearing and observation. Garman (1948) has taken the opposite view and he based his conception of the genus Typhlodromus on the ventral view alone, apparently ignoring the figures of the dorsum and the chelicerae. 2) In Southern Canada and in the adjacent northern states, by far the most common species on elm is a form which closely resembles Typhlodromus rhenanus Oudms. and which in all probability is Typhlodromus vepallidus (Koch).

8 8 H. H. J. NESBITT the type of the genus Typhlodromus. The exact identity of pyri and hence of the genus sensu stricto can never be established beyond all shadow a doubt. As first reviewer of the group, however, I believe that the evidence is sufficient to state that Typhlodromus pyri Sch., 1857, and T. tiliae Oudms., 1929 are very closely related if not identical, and that the specimens of T. tiliae in the Oudemans collection in the Rijksmuseum van Natuurlijke Historie, Leiden, bearing the labels "op Tilia platyphyllos Arnhem 1900" could be regarded as Typhlodromus pyri and considered illustrative of the genus Typhlodromus. In the event of these ever being lost there are specimens in my collection which were taken from pear trees from the Bonn area, Germany, which have been critically compared with Oudemans' material. The genus Iphidulus (type: I. communis Rib., ) which was established in 1902 by Ribaga is worthy of consideration. Vitzthum admitted it to his list of genera in the subfamily Phytoseiinae and Garman (1948) of type used the name for that group of species, which Oudemans placed in the genus Typhlodromus, wherein there are not more than seventeen pairs of dorsal setae and a large anal plate in the female (in other words the group which has just been discussed under the genus Typhlodromus sensu stricto). The question follows as to whether these two workers were justified in believing that this group had any validity. Three reasons are patent which seriously question their position: i) Ribaga's (1902) writings lead us to believe that he had no clear cut idea as to the limits or main characteristics of either the genus Seiulus or Iphidulus because while he stated in his "Gamasidi Planticoli" that the setae in the genus Seiulus are of the slightly denticulate plumiform type ("setis mediocribus, lenitre denticulato-plumatis armatum") and all those in the genus Iphidulus are simple not heavily pectinate or plumose ("omnibus autem simplicibus (nec phimosis)") he placed Gamasus vepallidus Koch in the genus Seiulus despite his assertion that it possesses simple setae ("setulis omnibus subtilibus"). Had he been consistent he would have placed G. vepallidus in his genus Iphidulus. Instead he chose a new 1) Ribaga's (1902) description of this genus and of Seiulus reads as follows in translation: "Genus Iphidulus Rib. n. gen. Body ovoid. Six posterior setae (caudal, precaudal, and caudoventral) longer than the scapular and genual setae, all others very small or non existant; all are simple not pectinated; posteriorly the female genital scutum is almost as wide as the anal scutum." "Genus Seiulus Berl. Body oval, with moderate setae armed with slightly denticulate feathers. Cephalolateral, humeral, scapular, infrascapular and median setae of unequal length, among themselves and [i.e., the setae named vary in length among themselves and are shorter than the caudal] scarcely shorter than the caudal."

9 A TAXONOMIC STUDY OF THE PHYTOSEIINAE 9 species, viz., I. communis as the type of the genus, ii) His (1902) descriptions both of his species and his genera are too vague to be of any value; they were not accompanied by any figures or ecological notes which might serve to elucidate the descriptions, and as far as can be ascertained, there are no type specimens for a final comparison, iii) The earlier use of the name Iphidulus by Berlese ( ) in setting forth his description of Laelaps (Iphidulus) vepallidus Koch (sive Laelaptis stabularis protonympha) could be considered sufficient to invalidate its use by another author and in a different context. As a result I am of the opinion that the resurrection of the genus Iphidulus from semiobscurity was unwarranted and that it would be better taxonomic practice if this name were placed on the list of dubious or unknown genera. On the basis of the above argument the question concerning Garman's use of the generic names Iphidulus and Typhlodromus can now be answered. It would appear that the only logical solution would be to transfer those mites which Garman described in the genus Iphidulus to the genus Typhlodromus, and to place the genus Iphidulus in the list of nomina dubia. The species remaining of those which Oudemans described in the genus Typhlodromus, viz., those with better than twenty-five pairs of dorsal setae and a small female anal plate, will have to be assigned to a new genus herein to be described. The genus Seiulus. Certain difficulties have also arisen in connection with the use of the name Seiulus which was introduced in 1887 by Berlese (vide , fasc. 41, N3) in his description of Seius (Seiulus) hirsutigenus and used again in the description of Seius (Seiulus) vepallidus (K.) Berl. ). Since x the term is commonly used in the literature pertinent to this group of Laelaptids, some attention should be given to its validity. In 1902 Oudemans stated that there is "reason enough to adopt the name Seiulus" to replace Berlese's genus Seius ), because of the incorrect use of the last mentioned 2 1) The subtitle of this description (vide A.M.S, fasc. 54 N8) reads "Seii obtusi nympha generans". 2) Koch originally erected a genus Seius in 1836 with Seius togatus as type. In 1842 he apparently forgot about this work and chose S. viduus as the type for another genus Seius. What S. viduus is no one knows but Berlese believed it to be Epicrius glaber. Thus Koch had two genera by the name of Seius one in 1836 and another in Only the former of these can stand. In 1881 Berlese erected a third genus Seius (the one in question above) and designated Seius echinatus Koch as the type. Since this name is preoccupied, an alternative one must be chosen and Oudemans (1902) has selected Seiulus Berl., Berlese (1921) overcame the difficulty himself by transferring the species described in his genus Seius to other genera.

10 10 H. H. J. NESBITT name by the Italian author, and in 1929 (vide supra) he still thought that the genus (Seiulus) should be retained for the more limited group of mites possessed of the large feather-like lateral setae. Berlese apparently took cognisance of Oudemans 1902 criticism because in 1921 he abandoned his, use of the genus Seius by transferring all the species previously attributed by him to that genus to other genera. In this synonymic work (1921) he recognized Ribaga's genus Echinoseius, type Seius (Seiulus) hirsutigenus Berl, and left but a single species, Seius (Seiulus) vepallidus K. (Berl.) *), in the genus Seiulus. As Berlese never gave a formal description of this genus, our knowledge of his conception must be based on his comments about the species which he first assigned to it, and as the two species placed in the genus are totally different (vide A.M.S. fasc. XLI, N3 and fasc. LIV, N8), the question of which is the type is important. Vitzthum (1941) has taken the attitude that since hirsutigenus (ibid., fasc. XLI, N3) was the first described it should be considered the type and on this basis he has placed Ribaga's genus Echinoseius in synonymy with Seiulus. Berlese (1921) on the other hand, in his Index Sinonimico designated Seius (Seiulus) vepallidus (Koch) as the type of his genus Seiulus ) and transferred S. hisutigenus to the 2 genus Ameroseius Berl., If this be accepted the genus Seiulus Berlese, 1921 will be based not on Koch's vepallidus but on what has been argued elsewhere in this paper, is an incorrect interpretation of this species and furthermore a form which has been so figured and described that it cannot be specifically identified or separated from the members of the established genus Typhlodromus. If, on the contrary, we accept Vitzthum's contention the genus Seiulus can stand on the type hirsutigenus. What hirsutigenus is again poses difficulties and two alternative explanations of its identity have been advanced one, that it is the immature stage of an already described species, and the other, that it is a valid species. In his original description Berlese (ibid., fasc. XLI, N3) hinted at the former, stating that it might be the protonymph of Seius hirsutus Koch (his quotation reads 'sive Seii hirsuti protonympha') and later, in the introduction to the "Mesostigmata", he stated 1) Some attention should be drawn to the apparent confusion which exists concerning the authority for this species. Oudemans (1929 and 1939) attributed the species vepallidus correctly to C. L. Koch, In 1887 Berlese ( ) cited Koch as the reference whereas in 1921 Kramer was given by Berlese. 2) Ribaga in 1902 had already done this but as has been pointed out elsewhere in this paper, his conception of the genus Seiulus and his statement of the characteristics of vepallidus are so at variance that I fail to see how he could assign the latter to the former.

11 A TAXONOMIC STUDY OF THE PHYTOSEIINAE II that S. hirsutigenus was the nympha generans of S. hirsutus. Subsequent work has in part confirmed his suspicions in that many now believe hirsutigenus to be the protonymph of either Seius echinatus Koch or Seius hirsutus Koch. If such be true, and I see no way of proving it in the absence of all type specimens, two results are evident: (i) on the basis of Berlese's statement in his revision of the genus Ameroseius, that S. echinatus ) and 1 S. hirsutus ) are identical, and taking into consideration Y itzthum's (1941) 1 r contention that these two species are synonymous with S. muricatus Koch, 1839 and Acarus corbicula Sowerby, 1806, it follows that A. corbicula on the basis of priority becomes the type of the genus Seiulus and (ii) that the genus Ameroseius Berlese, 1904 falls into synonymy with Seiulus because of the fact that both were erected on the same type (S. hirsutus). If on the contrary, we accept Vitzthum's (1941) position and regard hirsutigenus as being a valid species, it follows that it is the type of the genus Seiulus. Ribaga's Echinoseius (type hirsutigenus) falls into synonymy, and the genus Ameroseius stands on type A. corbicula. Since the evidence to my mind, for maintaining that hirsutigenus is the protonymph of S. hirsutus (= S. echinatus) is weak, I believe that we would be in a sounder position if we accepted the second of those two possible alternatives, which automatically rules out the possibility of Berlese's vepallidus being the type of Seiulus, and regarded hirsutigenus Berlese as a valid species and type of the genus Seiulus. This does not help our immediate problem much, but it does remove the genus Seiulus from the genera which have to be considered in this context. The only genus with which it might be confused is Phytoseius, but as the latter has only fourteen pairs of serrate setae and as the former has many more, this ought not to be possible. On the basis of hirsutigenus the genus Seiulus has in all probability a greater affinity with the genus Kleemannia, but until the types of both can be examined this must remain an open question. MORPHOLOGICAL NOTES The mites of the subfamily Phytoseiinae are, on the whole, small (300 to 600 ix long), generalized gamasids with few diagnostic features to distinguish them. The gnathosoma is typical of most Laelaptids. It may be divided into two portions, the distal which is composed dorsally of the epistome or tectum as Snodgrass (1948) calls it, and ventrally of the two halves of the hypostome or lower lip; and the proximal part which bears the palps and 1) Both species were described by Koch (1839) in Deut. Crust. Myr. Arachn. hirsutus in fasc. 24, t. 12, echinatus in fasc. 24, t. 13.

12 12 H. H. J. NESBITT is confluent with the body proper. The margin of the epistome which is roughly triangular or semi-hexagonal in shape may be entire or slightly serrate. In such species as Garmania bulbicola it is drawn out into little teeth at its lateral angles (giving it the appearance of a cat's jaw). Distally, the hypostome is divided into two halves or flap-like processes, each bearing four setae, by a groove through which the labrum may be seen. In this groove are the transverse rows of minute teeth which Oudemans referred to as the rima. (The number of rows varies in different species and the number of teeth per row). Proximally the hypostome is continuous with the basal ring of the gnathosoma. Beneath the hypostome, in ventral view, may be seen the triangular-shaped outer, and needle-like inner, coxal processes ), the 1 so-called maxillary malae or corniculae. Dorsad of these lie the chelate chelicerae. In the female these are simple unmodified shears with curved tips; in the male the moveable digit bears a copulatory apparatus or process which has been compared to a roe's horn. In some species both digits of the chelicerae are toothed, in others only the fixed. As in all Laelaptids the fixed digit also bears a small seta, the pilus dentilis, about one-third of the way from the tip; in some there is in addition to this a small peg-like structure near the joint of the moveable member and in still others laterally a flap-like process which is mildly denticulate along its free margin. The idiosoma of the adult is covered dorsally by a single dorsal shield (in the larvae there may be one large and three smaller shields), which varies in the degree or rugosity which it exhibits and in the number, nature and disposition of the setae which it bears. In this group of mites all types of dorsal shield may be found from the smooth, highly polished back of Amblyseius to the heavily sclerotized and deeply pitted dorsum of the mites of the genus Kleemannia. In the genus Typhlodromus the dorsum exhibits scale-like markings or imbrications in low relief. Somewhat the same condition is frequently visible in Garmania; in Kampimodromus the markings are still more distinct and in Phytoseius the ridges are in high enough relief to give the surface a mat appearance under the lower powers of magnification. The chaetotactic pattern likewise varies in different groups though in all the most constant feature is a distinct hexagonal area, on the anterior half of the shield, which is delimited on each side by setae D3, MT and D 2 4 ). For comparison, the genus Typhlodromus (vide text-fig. i) with its short setae of comparable length may be taken as the basic pattern. In Amblyseius the 1) Winkler (1886) states that the maxillary malae are processes of the hypostome. In the Laelaptidae there is reason for believing that the inner needle-like structures are processes of the labrum. 2) The terminology applicable to the setae has been adopted from Garman (1948).

13 A TAXONOMIC STUDY OF THE PHYTOSEIINAE 13 Text-figure 1. Typhlodromus massei n. sp., female. Left, dorsal surface, shewing L, lateral setae, D, dorsal setae, M, median setae and S, sacral setae. Right, ventral surface, shewing scuta and VLi, ventrolateral seta. Original. disposition of the setae is relatively the same but the setae themselves shew great differences in length; several of the marginal setae, viz., D x, L4, Le and Ls are very long (Ls equals almost one-half of the width of the body) and the median setae are minute. In the genera Kampimodromus and Phytoseius the chaetotactic pattern is recognizably similar to that of Typhlodro-

14 14 H. H. J. NESBITT mus but again the marginal setae are greatly modified in that they are slightly pectinated in the first-mentioned genus and heavily serrate in the second. In Garmania the basic pattern is altered by the addition of at least eight more pairs of setae but the nature of the setae is similar to that of Typhlodromus. Finally in Kleemannia both the disposition and nature of the setae is greatly modified. The ventral surface of the idiosoma is protected by a series of scuta or plates and bears the legs, the peritremata and the tritosternum. The latter, a forked structure of unknown but most probably sensory function, is composed of a short basal piece and two heavily pectinated flagella. It arises anterad of the sternal scutum and lies ventrad of the gnathosoma. The venter of the female is covered by several plates; the sternal is the most anterior of these and is roughly square in shape. In most of the species which will be discussed it bears three pairs of small bristles; in such forms as T. reticulatus, T. fallacis, and T. vitis, however, the rearmost pair of bristles stand on two small processes that extend posteriorly from the main body of the plate. In G. amboinensis, T. (N.) rhenanus, T. tiliae, T. tiliarum, and T. conspicuus, the third pair of bristles has moved posteriorly to arise from a pair of small platelets which may be either a logical outcome of the processes of isolation begun in such forms as T. reticulatus or a subdivision of the metapodal scuta which follow. These latter platelets each bear a bristle. Between the two metapodal platelets is an area which is mildly sclerotized. Most authors ) who have written about this group, regard this as an anterior 1 extension of the genital plate, the vulvar area through which the eggs pass; Garman (1948), however, calls it the ventral plate. It continues posteriorly as a well sclerotized genital plate which bears one pair of setae and is usually truncate posteriorly. Behind it lies the last of the ventral plates that called by Oudemans the ventrianal, by Garman the anal. In the majority of the species studied this is obviously a composite plate of hexagonal or rectangular shape made up of a small anal plate ) carrying three paraanal setae and a 2 larger anterior extension which bears an additional three or four pairs of preanal setae. In the remainder of the species, viz., those of the genus Garmania, only the small oval-shaped anal plate is present. In a few species such as T. tiliae the ventrianal scutum is separated from the genital plate by a narrow band of chitin in which lies a long narrow bar or a series of small bars of more heavily sclerotized chitin. The remainder of the ventral surface of the female is covered by weakly sclerotized chitin which extends dor sally 1) See Ewing (1929) for a somewhat different terminology. 2) The composite nature of this scutum is particularly evident in such forms as N. bakeri (Garm.) where the anal plate appears to be on top of the ventral plate.

15 A TAXONOMIC STUDY OF THE PHYTOSEIINAE 15 to fuse with the dorsal shield. In the systematic part of this paper it has been referred to as the interscutal membrane. Laterad of the ventrianal plate it may include one or two parapodial plates which have no great significance and laterad and posterad of coxa IV the peritremal plate which bears the stigma and continues forward as the peritrema. The interscutal supports a few setae: two just off membrane the dorsal shield, one in the humeral position Si and one in the sacral region So, three laterad of the ventrianal plate on each side, and a pair of ventro-lateral setae (VLi) near the posterior margin of the body. The ventral surface of the male differs quite radically from that of female in that it is on the whole more heavily sclerotized and is covered by only two large scuta. The anterior region between the coxae is paved with the sternal scutum which bears five setae which are homologous with those on the sternal, metapodal and genital scuta of the female. On its anterior margin may be seen the opening of the vas deferens and through the plate the vague outlines of the duct. The opisthosomal region behind coxae IV is covered by one large plate which usually bears five pairs of setae in addition to the three anal setae. As in the female there are three additional pairs of setae on the interscutal membrane, Si and S2 beside the dorsal shield and a pair behind the ventrianal plate. The legs of these mites bear few setae which have taxonomic significance. On the genu and tarsus of leg IV of many species are to be found modified setae which probably have a sensory function. KEY TO THE GENERA OF THE PHYTOSEIINAE Ai. All dorsal setae smooth, some (L4, 6, s) long and whip-like, others minute. B Bi. Ventrianal scutum of the female much wider than the genital; peritremal plate truncate Gen. Amblyseius Berlese*) B2. Ventrianal scutum of the female not conspicuously wider than the genital; peritremal plate blunt or approaching acute Subgen. Amblyseiopsis Garman 1 ) A2. Most dorsal setae smooth, some (such as those at the posterior margin) very faintly pectinate B Bi. With not more than 18 pairs of dorsal setae; ventrianal plate of female with at least three pairs of setae in addition to the paraanals ) 2 Gen. Typhlodromus Scheuten B2. With upwards of 25 pairs of dorsal setae; ventrianal plate of female may or may not have setae in addition to the paraanals C Ci. The moveable and fixed digits of the chelicerae of comparable length. D Di. Anal plate of female small and provided with only 3 setae Gen. Garmania n. gen. 1) The characters used for the separation of these two groups have been taken from Garman (1948). 2) In the keys "paraanal" is used as a collective term to include both the para- and postanal setae. the

16 16 H. H. J. NESBITT Do. Ventrianal plate of female larger and provided with 4 pairs of setae in addition to paraanals.... Subgen. Paragarmania n. subgen. C2. Fixed digit of chelicerae only one-third as long as the moveable and surmounted by a seta-like process Gen. Blattisocius Keegan A3. Many dorsal setae distinctly pectinate or serrate B Bi. Dorsum smooth or mildly imbricate, marginal setae noticeably pectinate, others may be Gen. Kampimodromus n. gen. B2. Dorsum distinctly rugose C Ci. Marginal setae long, serrate or thorny Gen. Phytoseius Rib. C2. All dorsal setae leaf-like and serrate.... Gen. Kleemannia Oudms. Genus Typhlodromus Scheuten Typhlodromus Scheuten, Archiv fur Naturges. 23: The genus Typhlodromus as understood in this paper is a closely-knit and coherent group of rapidly running predaceous mites. Its members may be recognized by the possession of the following characters: a weakly sclerotized body covered by a single dorsal shield which may be mildly imbricate and which bears approximately 17 pairs of simple rarely pectinate setae. The chaetotactic pattern of the dorsum (vide text-fig. 1) is basically made up of 6 dorsal, 2 median and 8 or 9 pairs of lateral setae; and in addition there are the scapular and lumbar setae which arise from the interscutal membrane laterad of the dorsal shield. The epistome may be smoothly rounded, sinuate or nearly pointed. It is never drawn out into a long spine. The chelicerae are denticulate with the arrangement of the teeth varying in different species. The male spermatophore is either T- or hammer-shaped. The sternal plate in the female usually bears three pairs of setae. Occasionally when there are only two pairs, the third and most posterior pair stands on a small pair of platelets located between the sternal plate in the metapodals. The ventrianal plate in the two sexes, whilst different in shape, bears three or four pairs of preanal setae in addition to the paraanals. The setae of the legs are of approximately the same length as those of the dorsum. The only exception to this being the tactile setae of leg IV which may on occasion be represented by one long whip-like or truncate seta on the tarsus or less frequently by a series of three setae on the last three members of the leg. Type of the genus: Typhlodromus pyri Scheuten (? T. tiliae Oudms.). It should be noted that Oudemans always included the two setae on either side of the anus (never, however, the postanal) in computing his figure of the number of setae on the ventrianal plate. In this paper the three paraanal setae (i.e., the two beside and one behind the anus) have been excluded from this figure as they are constant in all the Phytoseiinae.

17 A TAXONOMIC STUDY OF THE PHYTOSEIINAE 17 SYNOPSIS OF THE SPECIES AI. With 8 lateral setae B Bi. With setae L7, Ls, and M2 of equal length; ventrianal plate in both sexes provided with 4 pairs of preanal setae and a pair of pores T. conspicuus (Garman) B2. With seta L7 noticeably shorter than either M2 or Ls; ventrianal plate with 3 pairs of preanal setae but lacking a pair of pores C Ci. With Li equal in length to the distance between its base and that of D2 and with the tip of M2 extending well beyond the margin of the dorsal shield T. aberrans Oudms. C2. With Li equal to approximately one-half of the distance between its base and that of D2; tip of M2 does not extend beyond the margin of the dorsal shield T. vitis Oudms. A2. With 9 lateral setae B Bi. Seta M2 paired ) with one of either L7 or Ls x C Ci. Seta M2 paired with L7 D Di. The 6 setae of the dorsal hexagonal area shorter in length than the distance between their bases E Ei. Anterior lateral setae Li to L4 greater in length than the distance between their bases T. masseei n. sp. E2. Anterior lateral setae shorter by approximately one-half than the distance between their bases F Fi. Dorsum noticeably reticulate; ventrianal plate of female roughly triangular in shape and as broad as long T. reticulatus Oudms. F2. Dorsum only mildly imbricate; ventrianal plate hexagonal in shape and almost one and one-third times as long as broad T. cucumeris Oudms. D2. The setae of the dorsal hexagonal area are equal in length to the distance between their bases T. fallacis (Garman) C2. Seta M2 paired with Ls D Di. All dorsal setae exceedingly long; extend as far as or well beyond the bases of succeeding ones E Ei. Dorsal setae extend well beyond the bases of the succeeding; 4 pairs of preanal setae on ventrianal plate T. longipilus n. sp. E2. Dorsal setae extend as far as but not beyond the bases of the succeeding; only 3 pairs of preanal setae on the ventrianal plate T. occidentalis n. sp. D2. Dorsal setae shorter; tips do not extend beyond the bases of the succeeding T. tiliae Oudms, (=?T. pyri Scheuten) B2. Seta M2 either forming a triangle with two of the lateral setae or well removed from all other setae C Ci. Setae M2, L6 and L7 equidistant from each other and describing a triangle ; ventrianal plate of the female provided with 3 pairs of preanal setae situated on the anterior third of the plate... T. finlandicus Oudms. 1) When the expression paired is used in this paper it means that the bases of the two different setae referred to (M2 and one of the lateral group) are as close together as are the bases of the two setae D3. It might be noted that a line drawn through Lx M2 M2 Lx should be straight. The only point in the key where this definition is at all strained is the coupling Di of C2 of Bi of A2 where in the two species mentioned M2 and Ls are separated by a distance which is greater by one-quarter than the distance D3-D3.

18 18 H. H. J. NESBITT C2. Seta M2 unpaired with any of the lateral setae but equidistant from D5 and LT T. pomi (Parrott) A3. With 10 lateral setae B Bi. Setae M2, L5 and L6 form a triangular grouping T. (Neoseiulus) tiliacolus (Oudms.) B2. Setae M2, Ls and L9 form a triangular grouping C Ci. With 2 or 3 pairs of preanal setae on the ventrianal plate in both sexes. D Di. With only 2 pairs of preanal setae; ventrianal plate devoid of pores, in the female sole-shaped T. (Neoseiulus) soleiger (Ribaga) D2. With 3 pairs of preanal setae; ventrianal plate provided with a pair of pores, in the female hexagonal in shape T. (Neoseiulus) barkeri (Hughes) C2. With 4 pairs of preanal setae on the ventrianal plate D Di. Ventrianal plate simple but bearing a pair of crescent-shaped pores in addition to the setae T. (Neoseiulus) rhenanus Oudms. D2. Ventrianal plate modified to give the impression of an anal plate superimposed on the larger plate.. T. (Neoseiulus) bakeri (Garman) A4. With 11 lateral setae; M2 paired with L9 T. tiliarum Oudms. Typhlodromus tiliae Oudemans (? = T. pyri Scheuten) Plate IV; Plate V; Plate VIII figs. 1 and 3; Plate XI figs. 15 and 25. Typhlodromus tiliae Oudemans, Entom. Bericht. 8(169) : Typhlodromus tiliae Oudemans, ibid., 8(170) : 33 (additional notes). Typhlodromus tiliae Oudemans, ibid., 8(171) : 51 (redescription). In the first description alluded to above, Oudemans described Nympha I as follows (in translation) : "Nympha I: 191 fi long; width at the shoulders (flattened specimen) 120 n; oval in shape with the narrow end directed to the front. The dorsal shield bears 14 pairs of setae of which the two vertical, the two humeral, and the two at the rear margin are stronger than the others. Moreover, 8 setae are placed on the rear margin. The edge of the epistome is semicircular in shape. Tarsus IV is provided with a sensory seta which is one-half as long as the tarsus." In the third description Nympha I is characterized as follows: "As I now have a better example than formerly, I can now give a better description than previously. Length 235 fi; breadth below the shoulders 149 ; oval with the point towards the rear; in front of the shoulders slightly concave. Propodosomal shield with 9 pairs of setae; pygidial shield semicircular in shape with a rounded semi-circular extension of the anterior median border. It bears 4 pairs of setae of which two are longer and extend beyond the hind margin. In addition to these there are two intermediate shields in the interscutal membrane and 5 pairs of setae (2 at the shoulders). Anterad of the anal plate there are 3 pairs of bristles [and beside the plate] ) an additional pair that are a little thicker." 1 Female-' From the first description: "340 M long, width at the shoulders 190 V (flattened specimen), oval in shape with the 1) In this paper square [ ] brackets have been used to indicate an interpolation in the translation.